Stroke is a leading cause of death and disability and creates a societal burden of $34 billion each year in the United States alone. While new endovascular treatments have shown promise to improve stroke outcomes and recovery, they have been hampered by the lack of noninvasive biomarkers to identify patients who are good candidates for these therapies. In particular, imaging of brain oxygenation is a long-recognized but unmet need in the stroke community, and we propose to use magnetic resonance imaging (MRI) to address this need. This project develops a clinically feasible toolset to noninvasively image brain tissue oxygenation by magnetic resonance imaging. Our specific aims are (1) to develop new MRI techniques, quantitative BOLD and vascular fingerprinting, that utilize a novel pulse sequence to quantify oxygen extraction fraction (OEF) levels in cerebral tissues; (2) to validate MR OEF imaging with simultaneous [15O] positron emission tomography (PET) reference images in patients with cerebrovascular disease; and (3) to apply MRI OEF methods to study oxygenation status in patients with carotid artery stenosis who are at high risk of stroke. The innovation of this work lies in the development of a novel imaging technique to assess quantitative OEF in brain tissues. The use of a hybrid PET/MRI system to validate the new OEF imaging methods with simultaneous measurements by [15O] PET is also highly novel. The outcome of proposal is a validated MRI tool to image OEF, and quantitative MRI observations of OEF in patients with carotid artery stenosis, and its relationship to patient symptoms and perfusion status. The significance of this work is a clinically feasible MRI protocol to image tissue OEF that is transferrable to any clinical site. The reproducible and robust tool will allow the stroke imaging community to better stratify patients for new treatments, and implement multi-center trials to evaluate endovascular interventions to reduce stroke risk in these populations.